Drug addiction exacts a substantial toll on the economic, social, and medical systems in this country. The biology of brain reward systems mediating the reinforcing property of drugs is relatively well understood, but hedonic properties alone fail to explain why some individuals progress to uncontrolled drug taking and others do not. A complete understanding of addiction requires a shift in focus to negative reinforcement models in which a compulsion to take drugs is driven by the need to terminate or avoid an aversive state. The aversive state may be somatic or affective. Little is known about neural mechanisms of negative reinforcement. Animal studies suggest that negative learning involves not only opponent processes within the same reward-related systems but may also involve distinct neural structures and signals. It is clear that there are cognitive and behavioral impairments in motivational affect, inhibitory control, and frontal executive function in substance dependent individuals (SDI). The overall goal of this application is to elucidate mechanisms underlying negative motivational affective state and negative reinforcement learning deficits in SDI. This will be accomplished using functional MRI (fMRI) and a reinforcement learning paradigm based on the Iowa Gambling Task. We will show that the paradigm is sensitive to deficits in negative and positive reinforcement learning in SDI. We will study abstinent SDI and healthy controls matched for education, IQ, and age, to address 3 specific aims: Aim 1. Determine if there is a deficit in negative reinforcement learning in SDI compared to controls. If so, determine if this deficit is associated with a negative affective state. Aim 2. Determine the neural pathways involved in negative and positive reinforcement learning Aim 3. Determine if frontal-limbic activity differs in SDI compared to controls during negative and positive reinforcement learning